A neural networks-based numerical method for the generalized Caputo-type fractional differential equations

The paper presents a numerical technique based on neural networks for generalized Caputo-type fractional differential equations with and without delay. We employ the theory of functional connection-based approximation and the physics-informed neural network with extreme learning machine-based learning to solve the differential equation. The proposed method uses the L1 finite difference scheme and the Volterra integral equation scheme to create the loss function. The novelty of this work is the proposal of the neural network-based scheme coupling the idea of the theory of functional connections and a new loss function for the solution of generalized Caputo-type differential equations. The proposed approach is applied to single differential equations and the system of differential equations with single and multiple delays. & COPY; 2023 International Association for Mathematics and Computers in Simulation (IMACS). Published by Elsevier B.V. All rights reserved.